How does the way land is developed influence the way people travel? As regional population grows, can the transportation problems new people create be mitigated by changes to the density or style of the areas where they live or work? A new report of the Transportation and Regional Growth Study, Land Use and Travel Choices in the Twin Cities, 1958-1990, addresses these questions.

Authors Gary Barnes, formerly of CTS and now with the Humphrey Institute's State and Local Policy Program, and Professor Gary Davis of Civil Engineering analyzed the potentially hidden but important determinants of travel behavior. They found that, with some small exceptions, travel behavior was not very sensitive to the way land was developed in the Twin Cities area during 1958-1990.

Using 1990 data (the most recent available), the researchers updated an earlier analysis by Yacov Zahavi, who had studied changes during the period 1958-1970. They examined daily travel times for workers and non-workers, with worker time broken into commute time and other. They also examined whether people travel at all, the mode they use, and the number of trips they make.

What they found

Their primary result is support for Zahavi's earlier finding of a daily "budget" for time spent traveling. "People do not travel every single day, but on the days when they do, they spend on average about 70 minutes," the researchers say.

The second key finding is that while there are variations in some travel choices across the region, these differences arise more from variations in job access, job location, and population characteristics rather than from the way residential land is developed.

With the exception of commute times and mode choice (for example, bus or car), none of the ways of describing travel behavior showed variations greater than 10 percent within the metro region. The only way of dividing the population that yielded significant (more than 10 percent) differences in average daily total travel times is by location; travelers residing in outlying rural areas average 80 minutes a day, while central city travelers average about 68. "This finding contradicts the common belief that residents of dense urban neighborhoods spend much less time traveling each day than do suburban dwellers," Barnes says.

The differences in average travel times are due entirely to differences in commute times. Non-work travel was almost the same regardless of location. "People with plentiful shopping close to home spent just as much time on non-work travel as did residents of the most remote, purely residential suburban neighborhoods," the researchers say.

Commute times were determined by regional job access, not by the availability of jobs in the immediate vicinity of the home. That is, the availability of jobs within a mile of the home had almost no impact on average commute times, but the availability of jobs within a 20-minute drive did. Still, the large differences in job accessibility led to relatively small differences in commute times from one part of the urbanized area to another.

Only mode choice was influenced by land use. However, the impact was quite small compared to the size of the land use changes involved. An increase of 1,000 people per square mile is associated with at most about a one percent increase in transit share of work trips, and about a one percent decrease in overall daily driving time per person.

Although land use in the home neighborhood appears to have little impact on travel choices, land use at job destinations can have a fairly substantial effect on mode choice. About 30 of the 1,165 traffic zones in the region (less than one percent of the total land area) attract over 60 percent of all transit trips and a substantial percentage of walking and biking trips.

These zones—the central parts of the two downtowns, and the University of Minnesota campuses—attract average transit shares in excess of 15 percent even from remote suburban locations. "A possible explanation is that the very large size of these employment centers means that it is feasible to have frequent bus service, from all over the region. This frequent and fast service, combined with parking costs and inconvenience, makes transit a more attractive choice," Barnes and Davis say.

Overall, average daily travel times in 1990 were only about three percent higher than they were in 1958 (about two minutes), despite dramatic changes in land use and transportation infrastructure. The distance people travel varied considerably, but this variation was determined almost entirely by differences in average travel speed. That is, while higher speeds may have led to shorter times for individual trips, people used the time savings to make more or longer trips.

What they recommend

Barnes and Davis believe focusing policy on situations and people that behave differently from the norm could yield benefits. And while it may not be possible to reduce total auto travel by much, the negative effects of this travel could be tempered if attention is focused directly on the problems rather than on travel in general.

This led them to three main recommendations. First, policy should focus on "problem" travel rather than treating all travel equally. In this regard, work trips are key. They are far longer than any other type of trip, more congested, and because they are concentrated in time, they create the need for far more highway capacity than would be necessary otherwise.

Second, increasing the density of commercial development is much more likely to have a beneficial impact on travel behavior than is higher density in residential areas. Building on existing success might be an effective way of integrating transit use into the regional way of life. "For example, people create temporary density at park-and-ride bus stops. It may be possible to provide good bus service to high-density suburban job centers surrounded by low-density housing," the researchers say.

Finally, it might be possible to contain the growth of auto use by helping the minority of residents--such as downtown workers and students—who desire a less auto-intensive lifestyle. Creating opportunities for as many people as possible to live in those neighborhoods close to high-density commercial or educational areas, where residents more often use modes other than autos, might make sense.

Complete report available

The report is available in printed form or download as a PDF file. PDF viewer software such as the free Adobe Acrobat Viewer (www.adobe.com/acrobat) is required to view PDF documents.

Barnes, Gary, and Gary Davis. Land Use and Travel Choices in the Twin Cities, 1958–1990. Report No. 6 in the series: Transportation and Regional Growth. Minneapolis: Center for Transportation Studies, 2001.
Download PDF document (1.14MB)

A limited number of printed reports and CD-ROMs are also available by calling the Center at 612-626-1077 or e-mailing CTS@tc.umn.edu.